IMO: Greedy Layer-Wise Sparse Representation Learning for Out-of-Distribution Text Classification with Pre-trained Models
This work addresses the challenge of out-of-distribution text classification for machine learning models, representing an incremental advancement in domain-specific generalization techniques.
The paper tackled the problem of domain generalization in text classification by proposing IMO, a method that learns invariant features through sparse mask layers and token-level attention, resulting in substantial performance improvements over strong baselines across various metrics and settings.
Machine learning models have made incredible progress, but they still struggle when applied to examples from unseen domains. This study focuses on a specific problem of domain generalization, where a model is trained on one source domain and tested on multiple target domains that are unseen during training. We propose IMO: Invariant features Masks for Out-of-Distribution text classification, to achieve OOD generalization by learning invariant features. During training, IMO would learn sparse mask layers to remove irrelevant features for prediction, where the remaining features keep invariant. Additionally, IMO has an attention module at the token level to focus on tokens that are useful for prediction. Our comprehensive experiments show that IMO substantially outperforms strong baselines in terms of various evaluation metrics and settings.